I too was intrigued by Walker’s suggestion that the non-linearity that ABW exhibits was intended to compensate for poor quality screens. I’ve posted a linearity checker plot for ABW on EEM here on IJM before and there’s a significant contrast boost compared to linear that looks to be intentional, so that fits. However the same plot for a semi-gloss paper shows ABW to be much lighter across the range, which I don’t think fits, as most screens are set too bright for print. I don’t understand the rationale at all.
[QUOTE=LarryB;10406]Brian has generously taken my images and processed them to make equivalently toned ABW and piezography prints for me to compare. Apparently he’s done a couple of “tricks” to produce a more linear print using ABW, I think. I’ll have to reread his e-mail, but I think it involves saving the file with a profile that offsets the non-linearity of the ABW print.[/QUOTE]
To be clear, I’ve taken the image that Larry is using as a test image, and printed it using Piezography (P2/K6), QTR-K3-OEM and ABW. I’ve done this on a matte paper and a semi-gloss. The P2 & QTR-K3-OEM prints are tonality-matched by virtue of the fact that the QTR curves are linear. The ABW prints are different because it’s not linear.
To enable Larry to make an apples-with-apples comparison of all three B&W printing approaches, I’ve reprinted the ABW version with a curve applied in Photoshop that corrects for the non-linearity of ABW, so that its overall tonality matches the other two prints, and so you can compare it more precisely to P2/K6 & QTR-K3-OEM. [I have a spreadsheet that takes the numbers in a linearisation check, inverts them, expands the range from (dmin,dmax) to (0,100) and gives me the points to use in a Photoshop curve. A technique I sometimes used before the advent of the relinearisation droplet.]
I’ve also printed the image using P2/K6 with a Photoshop curve applied that mimics the non-linearity of ABW, as another way to get the two printing approaches closer, to enable a fairer comparison.
The matched prints are not exact, as despite my best efforts and my spreadsheet and my relinearised QTR curves, it’s still an approximate match, and the three printing approaches render neutral quite differently, ABW being noticeably cool. But I think they’re as close as they can reasonably be made to be, and hopefully will assist Larry to make an informed choice.